9:50 p.m.
Hello,
I need to take the log difference of a matrix, i.e. log(M[2 rows(M):,]/M[1:rows(M)-1,]).
Unfortunately, M has elements equal to zero. I need to replace the nan's and inf's
with 0's. This almost works
M = isnan(M) ? 0 : M
but does not remove inf's. Any sugestions?
9:28 a.m.
On Tue, 23 Jul 2013, Logan Kelly wrote:
I need to take the log difference of a matrix, i.e. log(M[2
rows(M):,]/M[1:rows(M)-1,]). Unfortunately, M has elements
equal to zero. I need to replace the nan's and inf's with
0's. This almost works
M = isnan(M) ? 0 : M
but does not remove inf's. Any sugestions?
Is this calculation actually legit? Assuming it is, then
<hansl>
matrix M = muniform(15,2)
M[2,2] = 0
M[3,2] = 0
matrix ldM = log(M[2:rows(M),] - M[1:rows(M)-1,])
ldM = isnan(0 * ldM) ? 0 : ldM
print M ldM
</hansl>
This relies on the IEEE 754 rules: both 0*inf and 0*(-inf)
return nan.
Allin Cottrell
10:56 a.m.
This is great! Actually, I was looking for something like this for a
while. Could anybody explain the single elements of this command,
please? It does not look very intuitive to me. Or is there reference in
the gretl guide on this?
M = isnan(M) ? 0 : M
Thank you.
Artur
Am 25.07.2013 11:28, schrieb Allin Cottrell:
On Tue, 23 Jul 2013, Logan Kelly wrote:
> I need to take the log difference of a matrix, i.e. log(M[2
> rows(M):,]/M[1:rows(M)-1,]). Unfortunately, M has elements
> equal to zero. I need to replace the nan's and inf's with
> 0's. This almost works
>
> M = isnan(M) ? 0 : M
>
> but does not remove inf's. Any sugestions?
Is this calculation actually legit? Assuming it is, then
<hansl>
matrix M = muniform(15,2)
M[2,2] = 0
M[3,2] = 0
matrix ldM = log(M[2:rows(M),] - M[1:rows(M)-1,])
ldM = isnan(0 * ldM) ? 0 : ldM
print M ldM
</hansl>
This relies on the IEEE 754 rules: both 0*inf and 0*(-inf)
return nan.
Allin Cottrell
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
--
Jabber account: t.artur(a)ubuntu-jabber.de
Join me on Diaspora: ArturT(a)kosmospora.net
(Invitation to Diaspora: kosmospora.net/i/94ce021fe957
You would like to send an encrypted PGP message?
-----BEGIN PGP PUBLIC KEY BLOCK-----
Version: GnuPG v1.4.12 (GNU/Linux)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=H2AF
-----END PGP PUBLIC KEY BLOCK-----
11:48 a.m.
It is an if condition on the single elements of a series
if (isnan(M[i]) = 1, i.e. the correspondent element of M is null) then
M[i] = 0 else
M[i] = M[i].
Bye
Giuseppe
On Wed, 2013-07-31 at 12:56 +0200, Artur T. wrote:
This is great! Actually, I was looking for something like this for a
while. Could anybody explain the single elements of this command,
please? It does not look very intuitive to me. Or is there reference in
the gretl guide on this?
M = isnan(M) ? 0 : M
Thank you.
Artur
Am 25.07.2013 11:28, schrieb Allin Cottrell:
> On Tue, 23 Jul 2013, Logan Kelly wrote:
>
>> I need to take the log difference of a matrix, i.e. log(M[2
>> rows(M):,]/M[1:rows(M)-1,]). Unfortunately, M has elements
>> equal to zero. I need to replace the nan's and inf's with
>> 0's. This almost works
>>
>> M = isnan(M) ? 0 : M
>>
>> but does not remove inf's. Any sugestions?
>
> Is this calculation actually legit? Assuming it is, then
>
> <hansl>
> matrix M = muniform(15,2)
> M[2,2] = 0
> M[3,2] = 0
> matrix ldM = log(M[2:rows(M),] - M[1:rows(M)-1,])
> ldM = isnan(0 * ldM) ? 0 : ldM
> print M ldM
> </hansl>
>
> This relies on the IEEE 754 rules: both 0*inf and 0*(-inf)
> return nan.
>
> Allin Cottrell
> _______________________________________________
> Gretl-users mailing list
> Gretl-users(a)lists.wfu.edu
> http://lists.wfu.edu/mailman/listinfo/gretl-users
>
12:40 p.m.
Thank you Giuseppe. As I said, very useful to know.
Cheers,
Artur
Am 31.07.2013 13:48, schrieb Giuseppe Vittucci:
It is an if condition on the single elements of a series
if (isnan(M[i]) = 1, i.e. the correspondent element of M is null) then
M[i] = 0 else
M[i] = M[i].
Bye
Giuseppe
On Wed, 2013-07-31 at 12:56 +0200, Artur T. wrote:
> This is great! Actually, I was looking for something like this for a
> while. Could anybody explain the single elements of this command,
> please? It does not look very intuitive to me. Or is there reference in
> the gretl guide on this?
>
> M = isnan(M) ? 0 : M
>
> Thank you.
> Artur
>
> Am 25.07.2013 11:28, schrieb Allin Cottrell:
>> On Tue, 23 Jul 2013, Logan Kelly wrote:
>>
>>> I need to take the log difference of a matrix, i.e. log(M[2
>>> rows(M):,]/M[1:rows(M)-1,]). Unfortunately, M has elements
>>> equal to zero. I need to replace the nan's and inf's with
>>> 0's. This almost works
>>>
>>> M = isnan(M) ? 0 : M
>>>
>>> but does not remove inf's. Any sugestions?
>>
>> Is this calculation actually legit? Assuming it is, then
>>
>> <hansl>
>> matrix M = muniform(15,2)
>> M[2,2] = 0
>> M[3,2] = 0
>> matrix ldM = log(M[2:rows(M),] - M[1:rows(M)-1,])
>> ldM = isnan(0 * ldM) ? 0 : ldM
>> print M ldM
>> </hansl>
>>
>> This relies on the IEEE 754 rules: both 0*inf and 0*(-inf)
>> return nan.
>>
>> Allin Cottrell
>> _______________________________________________
>> Gretl-users mailing list
>> Gretl-users(a)lists.wfu.edu
>> http://lists.wfu.edu/mailman/listinfo/gretl-users
>>
>
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
--
Jabber account: t.artur(a)ubuntu-jabber.de
Join me on Diaspora: ArturT(a)kosmospora.net
(Invitation to Diaspora: kosmospora.net/i/94ce021fe957
You would like to send an encrypted PGP message?
-----BEGIN PGP PUBLIC KEY BLOCK-----
Version: GnuPG v1.4.12 (GNU/Linux)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=H2AF
-----END PGP PUBLIC KEY BLOCK-----
4:54 a.m.
New subject: Random draw from a known discrete distribution
Dear all,
I am running gretl 2016b on Linux Mint Debian 2.
I would like to sample from a known discrete distribution.
In particular, given a vector of probabilities p = (p_1,p_2...p_n)
(where n can change) with associated natural numbers i = (1,2...n), I
would like to make a draw from this distribution.
So, my first question is: is there in gretl a built-in function to do
this?
Of course I can write down some code to accomplish the task, but in R
the function "sample" can directly do it.
So, in R the simple function:
sampleDist = function(p) {
sample(seq(1,length(p)),1,replace=T,p)
}
works as expected, where p is a vector of probabilities.
Example:
p <- c(0.1, 0.4, 0.2, 0.3, 0.0)
sampleDist(p)
Therefore, I tried to use R shared libraries and to define functions in
R in a foreign block in gretl
set R_functions on
foreign language=R
sampleDist = function(p) {
sample(seq(1,length(p)),1,replace=T,p)
}
end foreign
The code is executed in gretl and returns no error, but, when I recall
the function, gretl returns an error of "unknown function"
? matrix p = {0.1,0.1,0.5,0.3,0}
Generated matrix p
? scalar c = R.sampleDist(p)
c = R.sampleDist(
The symbol 'R.sampleDist' is
undefined
Error executing script: halting
This is an issue that I encounter also when I try to use some simple
built-in R function from gretl.
For instance, this code (taken from the guide) does not work:
? set R_functions on
? scalar b=R.choose(10,4)
b=R.choose(
The symbol 'R.choose' is undefined
Error executing script: halting
My R version is 3.1.1 and has been built with the shared library enabled
(this is the default option in the Debian package for R) and I checked
that gretl has correctly detected the path to R library libR.so in my
pc.
Is this a problem with the R version and I should upgrade/downgrade R?
Any idea or suggestion?
Best,
Giuseppe
--
Giuseppe Vittucci Marzetti
Assistant Professor, PhD
Department of Sociology and Social Research
University of Milano-Bicocca
Via Bicocca degli Arcimboldi 8 - Building U7 - Floor III - Room 330
20126 Milan, Italy
Phone: +39 0461 882162
Fax: +39 0461 882222
E-mail: giuseppe.vittucci(a)unimib.it
8:11 a.m.
New subject: Random draw from a known discrete distribution
On Sat, 11 Jun 2016, Giuseppe Vittucci wrote:
Dear all,
I am running gretl 2016b on Linux Mint Debian 2.
I would like to sample from a known discrete distribution.
How do you like this?
<hansl>
function matrix msample(matrix p, scalar r, scalar c)
ret = {}
matrix p = vec(p) # ensure p is a column
matrix p /= sumc(p) # and it's a probability vector
if minc(p) >= 0
ret=sumr(muniform(r*c, 1) .< cum(p)')
endif
return ret
end function
nulldata 200
set seed 123
series x = msample({0.1,0.6,0.3},$nobs,1)
freq x --plot=none
</hansl>
Perhaps we could make this an internal function. Allin?
-------------------------------------------------------
Riccardo (Jack) Lucchetti
Dipartimento di Scienze Economiche e Sociali (DiSES)
Università Politecnica delle Marche
(formerly known as Università di Ancona)
r.lucchetti(a)univpm.it
http://www2.econ.univpm.it/servizi/hpp/lucchetti
-------------------------------------------------------
8:46 a.m.
New subject: Random draw from a known discrete distribution
On Sat, 11 Jun 2016, Riccardo (Jack) Lucchetti wrote:
On Sat, 11 Jun 2016, Giuseppe Vittucci wrote:
> Dear all,
>
> I am running gretl 2016b on Linux Mint Debian 2.
> I would like to sample from a known discrete distribution.
How do you like this?
<hansl>
function matrix msample(matrix p, scalar r, scalar c)
ret = {}
matrix p = vec(p) # ensure p is a column
matrix p /= sumc(p) # and it's a probability vector
if minc(p) >= 0
ret=sumr(muniform(r*c, 1) .< cum(p)')
endif
return ret
end function
</hansl>
Whoops, forgot one line, sorry!
<hansl>
function matrix msample(matrix p, scalar r, scalar c)
ret = {}
matrix p = vec(p) # ensure p is a column
matrix p /= sumc(p) # and it's a probability vector
if minc(p) >= 0
ret = sumr(muniform(r*c, 1) .< cum(p)')
ret = mshape(ret, r, c)
endif
return ret
end function
</hansl>
-------------------------------------------------------
Riccardo (Jack) Lucchetti
Dipartimento di Scienze Economiche e Sociali (DiSES)
Università Politecnica delle Marche
(formerly known as Università di Ancona)
r.lucchetti(a)univpm.it
http://www2.econ.univpm.it/servizi/hpp/lucchetti
-------------------------------------------------------
9:12 a.m.
New subject: Random draw from a known discrete distribution
On Sat, 11 Jun 2016, Riccardo (Jack) Lucchetti wrote:
On Sat, 11 Jun 2016, Riccardo (Jack) Lucchetti wrote:
> On Sat, 11 Jun 2016, Giuseppe Vittucci wrote:
>
>> Dear all,
>>
>> I am running gretl 2016b on Linux Mint Debian 2.
>> I would like to sample from a known discrete distribution.
>
> How do you like this?
>
> <hansl>
> function matrix msample(matrix p, scalar r, scalar c)
> ret = {}
> matrix p = vec(p) # ensure p is a column
> matrix p /= sumc(p) # and it's a probability vector
> if minc(p) >= 0
> ret=sumr(muniform(r*c, 1) .< cum(p)')
> endif
> return ret
> end function
> </hansl>
Whoops, forgot one line, sorry!
<hansl>
function matrix msample(matrix p, scalar r, scalar c)
ret = {}
matrix p = vec(p) # ensure p is a column
matrix p /= sumc(p) # and it's a probability vector
if minc(p) >= 0
ret = sumr(muniform(r*c, 1) .< cum(p)')
ret = mshape(ret, r, c)
endif
return ret
end function
</hansl>
OMG, today's not my day, is it?
ret = sumr(muniform(r*c, 1) .< cum(p)')
should be
ret = sumr(muniform(r*c, 1) .> cum(p)')
Sorry for the noise, people.
-------------------------------------------------------
Riccardo (Jack) Lucchetti
Dipartimento di Scienze Economiche e Sociali (DiSES)
Università Politecnica delle Marche
(formerly known as Università di Ancona)
r.lucchetti(a)univpm.it
http://www2.econ.univpm.it/servizi/hpp/lucchetti
-------------------------------------------------------
2869
days inactive
3923
days old
8 comments
5 participants
participants (5)
-
Allin Cottrell -
Artur T. -
Giuseppe Vittucci -
Logan Kelly -
Riccardo (Jack) Lucchetti